Information Technology University (ITU) congratulates Mehwish Ghafoor on the successful defense of her Ph.D. thesis titled “Estimation of Human Pose Under Occlusion Using Deep Learning Techniques.”
Supervised by Prof. Dr. Arif Mahmood, her research addresses one of the key challenges in computer vision — accurately estimating 3D human poses even when body parts are partially or fully occluded. By introducing advanced deep learning frameworks such as Occlusion-Guided Temporal Dilated CNN (T3D-CNN) and Dual Transformer Fusion (DTF), the study enhances pose estimation accuracy and action recognition performance under real-world constraints.
The GEC included Dr. Mohsen Ali, Dr. Ijaz Akhter, and Dr. Waqas Sultani, while Dr. Murtaza Taj and Dr. Shahid Fareed served as local evaluators.
ITU extends warm congratulations to Mehwish Ghafoor and her supervisor on this remarkable academic achievement.